Bedini Ferris Wheel Regauging Motor

[Les_K]

Hi Les,

I just finished watching #28 and printed off the circuit that Ron posted earlier in this thread.

It is not clear to me that the mod adding N-channel mosfets is correct. At present, I think that the mosfets are leaching energy from the battery continuously.

I stumbled across this problem myself just a while ago. N-channel mosfets have an internal diode from source (anode) to drain (cathode) which appears to allow current to flow continuously if you hook the drain to the battery -ve.

Here is a simulation of the simple circuit that I ran into trouble with:

simple lamp circuit

and here is a simulation of important bits of your circuit:

abridged pulser

I thought that I saw JB draw p-channel fets on the video, but I need to view it again.

pt

Paul, I just went back through my notes from the conference and I think you got it!
my notes do show P channel. So unless I wrote it down wrong you just solved my major hurdle. I don't have any P-channel so looks like I'll be a few weeks out.

Thank's
Les

[minoly]

With no other instruction, I just decided to start adding different air coils. I should have done this long ago. if anyone is still having problems getting more out of the 3GT or even just running a single coil off of the Bedini Cole switch on a monopole wheel - just start trying out different coils in parallel with the main power coil but away from the wheel, add a core, remove the core - it does not take long to get the hang of the tuning.

Patrick

[hherby]

Paul, I just went back through my notes from the conference and I think you got it!
my notes do show P channel. So unless I wrote it down wrong you just solved my major hurdle. I don't have any P-channel so looks like I'll be a few weeks out.

Thank's
Les

I reviewed my pictures from the conference and JB was using a pair of IRFP260's in the comparator cap dumper.

Comparator was an LM741CN

Alex

[minoly]

Did I miss a post some where? 3 times per second still sounds too fast.

It was mentioned here several times and also on the video series.


Les

Les,
I have spent some time this morning reviewing the DVD series as well as searching this forum. please excuse my lameness, but I can not find where JB references 3+/- times per second.

My interest is in reading/listening to JB's actual words. Can anyone point me to the post or DVD - we have them all.

Our build seems to get more out of high amp pulse/low impedance and not so much about the freq. any time I increase the freq thereby lowering the amps in the pulse, charging suffers.

Thanks,
Patrick

[pault]

I reviewed my pictures from the conference and JB was using a pair of IRFP260's in the comparator cap dumper.

Comparator was an LM741CN

Alex

IRFP260 is an N-channel, with the S-D diode. I believe that JB said to flip the S and D, so, at a minimum, that's wrong in the diagram. I also think that the diagram shows + and - backwards for U1 (compared to vid #28). I'll continue to to review...

I think that I "get" what the design intent is, regarding the dump on negative. The cap gets an ever-increasing voltage and gets hit by 300+V spikes. The circuitry "wants" a stable operating voltage and needs to be protected from the spikes. The circuit common, then, is chosen to be the cap -ve and the zeners provide a stable voltage +5 and +10 (+15) "above" the cap -ve, regardless of what's happening on the cap +ve side. When the circuit common reference is cap -ve, then the mosfets probably need to be referenced to cap -ve also (esp. S to G threshold voltage).

More pondering is in order...

pt

[RS_]

Hi All,
For the last 3 weeks we have been testing a 2n4401 based Bedini Fig 33 SG self oscillator with a Conference Comparator Triggered cap pulser battery charger, running from 1.7V up to 12V or so. While charging 5ah 12V SLA. In the process of trying to get the Comparator circuitry's current draw as low as possible and still function correctly, I too found that current was flowing from the charging battery if the storage cap was not being charged, and suspected the internal mosfet diodes, so i tried a BD243C that does not have such diode, and it too let current flow
No part in place = No current flow

RS

[BrentA929]

Bedini GT3

Hey Guys,

Happy New Years to everyone! I have been following the most recent activity, but don't have much to add. Other than really wanting to get this comparator circuit figured out.

As far as FETs, I would have to agree with Alex. IRFP260s are what I have in my notes/pictures from the conference.

**The thing that bothers me the most about this whole comparator circuit thing. This is not what we were shown at the 2010 Conference. We were told "one black wire in, one black wire out." And this was pretty clear on the pictures and at the conference. Supposedly easy to do with a FET and a Zener. Ha! Several discussions were had about this in previous posts. Now we are hot on the trail of this new circuit. What is right in front of me IS very confusing!!! It's confused in my head. Or at least that was what I was told by JB when I asked him about this at the 2011 Conference. What happened to the "one black wire in, one black wire out?"

Sure makes it hard to follow...


Thanks, Brent


@Patrick,

Here is a link to one place JB mentions dump frequency...post 155

Les,
I have spent some time this morning reviewing the DVD series as well as searching this forum. please excuse my lameness, but I can not find where JB references 3+/- times per second.

My interest is in reading/listening to JB's actual words. Can anyone point me to the post or DVD - we have them all.

Our build seems to get more out of high amp pulse/low impedance and not so much about the freq. any time I increase the freq thereby lowering the amps in the pulse, charging suffers.

Thanks,
Patrick

[minoly]

Hey Guys,




Thanks, Brent


@Patrick,

Here is a link to one place JB mentions dump frequency...post 155

Thanks Brent,
I think that's a good goal for all of you that are building to spec - 13 footers with the same cap and coils JB uses.
for us little guys with little wheels and smaller coils. I think we have to adjust expectations. I notice on the most recent DVD28 both wheels are dumping about 1/second which is about how often mine dumps using a 3300uf 350volt cap at 48volts.

Thanks again - shows how much I forget - saw a show this week on a study that shows dementia setting in at 45 years old - time to get serious about doing cross word puzzles LOL :-)

Patrick

[pault]

more comparator observations

In the video, JB said he uses the "reference voltage" (5.1 zener, also attached to U1) to drive the collector of the H11D1 (pin 5). Battery -ve is connected only to the drain(s) of the FET, not to the collector(s) of the NPN's. In fact, he doesn't draw in the extra darlington MPSA06.

At the moment, I can see how to put this circuit in a box using 3 wires, but not 2.

1. cap neg, 2. cap pos, 3. battery neg (assuming bat pos is connected to cap pos outside of the box).

pt

[BrentA929]

Bedini GT3

pt,

Please don't let my frustration confuse you...


At the 2010 Conference = "one wire in, one wire out"
At the 2011 Conference = Comparator circuit


Not sure why it changed, but it confused me and I didn't get a very clear answer from JB.

Oh well, maybe at the 2012 Conference = Answers


Best Regards, Brent

In the video, JB said he uses the "reference voltage" (5.1 zener, also attached to U1) to drive the collector of the H11D1 (pin 5). Battery -ve is connected only to the drain(s) of the FET, not to the collector(s) of the NPN's. In fact, he doesn't draw in the extra darlington MPSA06.

At the moment, I can see how to put this circuit in a box using 3 wires, but not 2.

1. cap neg, 2. cap pos, 3. battery neg (assuming bat pos is connected to cap pos outside of the box).

pt

[Les_K]

In the video, JB said he uses the "reference voltage" (5.1 zener, also attached to U1) to drive the collector of the H11D1 (pin 5). Battery -ve is connected only to the drain(s) of the FET, not to the collector(s) of the NPN's. In fact, he doesn't draw in the extra darlington MPSA06.

At the moment, I can see how to put this circuit in a box using 3 wires, but not 2.

1. cap neg, 2. cap pos, 3. battery neg (assuming bat pos is connected to cap pos outside of the box).

pt

Wow I was gone for a few days and wham lots of stuff! Just got caught up so sorry for not responding earlier. I don't have the video yet so I can't speak to that. I do have my notes from the conference and I tried to draw as precisely as I could. And I do have a Darlington drawn driving the fet. it shows the fet connected directly to the batt neg. However I made the choice of an MPSA06. The circuit posted was one that I had tested and I knew would work well enough to get started. I was shooting for the lowest power usage I could. after playing with this Darlington setup I could drop the current to the H11D1 led by adding a much higher resistance and still get the switching. I am wondering now if that might be part of the problem. I wonder if the H11D1 is not fully switching because R9 is to high? Anyway I hope we can get some real testing in so we don't get caught up in guessing. and just remember what John said about the rock hitting you in the eye.........

Way to go
Les

[Les_K]

pt,

Please don't let my frustration confuse you...


At the 2010 Conference = "one wire in, one wire out"
At the 2011 Conference = Comparator circuit


Not sure why it changed, but it confused me and I didn't get a very clear answer from JB.

Oh well, maybe at the 2012 Conference = Answers


Best Regards, Brent

Hi Brent,
Good to hear from you again. I think I understand what you are talking about. but I am pretty sure John will never be able to discuss that original circuit.

Les

[pault]

[headsmack]I coulda had a V8[/headsmack]

I now think that this can be done with one wire in one out.

I began reviewing this thread from near-beginning, hit nvisser's post #185 page 7 and it triggered an (as yet untested) idea.

In LK/RC's diagram (and the video), the power and reference voltages are taken from the cap pos.

Instead, unhook D1 and R4 from cap pos and hook them to bat neg (and readjust the ref voltage).

Now, the whole circuit runs from one wire "in" (bat neg) and one wire "out" (cap neg).

Bat neg "rides" 12-16V (depending on SOC) below the voltage at cap pos. E.g. when the cap is at +50V, bat neg is at 50V-12V = +38V. You just have to make sure that you leave enough headroom to power the op-amp (15V for 741, 10V for TLV3702IP, etc).

I'll play with this when I have time and report.

Q: Is there an advantage/importance to having only one wire in and one out?

pt

[Les_K]

I now think that this can be done with one wire in one out.

I began reviewing this thread from near-beginning, hit nvisser's post #185 page 7 and it triggered an (as yet untested) idea.

In LK/RC's diagram (and the video), the power and reference voltages are taken from the cap pos.

Instead, unhook D1 and R4 from cap pos and hook them to bat neg (and readjust the ref voltage).

Now, the whole circuit runs from one wire "in" (bat neg) and one wire "out" (cap neg).

Bat neg "rides" 12-16V (depending on SOC) below the voltage at cap pos. E.g. when the cap is at +50V, bat neg is at 50V-12V = +38V. You just have to make sure that you leave enough headroom to power the op-amp (15V for 741, 10V for TLV3702IP, etc).

I'll play with this when I have time and report.

Q: Is there an advantage/importance to having only one wire in and one out?

pt

Paul,
I think you may have done it again....
this may take a minute to explain but this was what my testing resulted in today and I think it directly applies to the thought you just presented. I reduced the resistor R9 to 5.6 k. (by the way I have a LED between it and ground so I can see the pulses.). Now I was charging a little tractor battery and the voltage was 12.99 volts. I changed the resistor and after about an hour it was down to about 12.83 volts.

I got to thinking about this and realized what you had said about these changes. and I realized what might be happenning. My little test runner is only pulling a total of 370ma. So to run a full pulser circuit I am stealing far to much power to get any decent charging. I increased the resistor and charging got better. I think this is why the SCR circuit under these conditions is so much better. My test unit is marginal.

But here is a thought. The way mine is wired I think it might be stealing some power from the battery under charge. if so that's bad. We will make them batteries into Maracas..... Thus the value of your comment above.

One last thing I noticed in a comment from the conference that had to do with reaching the cap peak. My charge is a linear rise with no curvature. I am using 450volt caps and I think I should be using more like 50-75volt caps.

Anyway I think your right we need to get it running solely on the cap. So more testing is in order.
and Brent, thanks for bringing this up.
Les

[BrentA929]

Comparator circuit

Hey Les,

One other thing that you may try...

Based on the schematic that Ron posted back in post 1629, which he and I worked on together. You may try adjusting the capacitors C1 and C2. What I have found is that is where some of the biggest differences are found. I currently have 33uF for both of those. Seems to work pretty good.

Someone with a good scope will have to confirm this since the scope I have isn't very good. But what I can see visually is that the higher capacitance on C1 intensifies the brightness of the dump LED and seems to dump harder and the higher the capacitance of C2 the lower the main cap dumps in voltage. I guess it allows more time for the main cap to dump more!

Give it a try and let me know your thoughts.


Thanks, Brent


@pt,

Not sure of any benefit to the "one wire in, one wire out," merely an observation from the 2010 conference to the 2011 conference.

Paul,
I think you may have done it again....
this may take a minute to explain but this was what my testing resulted in today and I think it directly applies to the thought you just presented. I reduced the resistor R9 to 5.6 k. (by the way I have a LED between it and ground so I can see the pulses.). Now I was charging a little tractor battery and the voltage was 12.99 volts. I changed the resistor and after about an hour it was down to about 12.83 volts.

I got to thinking about this and realized what you had said about these changes. and I realized what might be happenning. My little test runner is only pulling a total of 370ma. So to run a full pulser circuit I am stealing far to much power to get any decent charging. I increased the resistor and charging got better. I think this is why the SCR circuit under these conditions is so much better. My test unit is marginal.

But here is a thought. The way mine is wired I think it might be stealing some power from the battery under charge. if so that's bad. We will make them batteries into Maracas..... Thus the value of your comment above.

One last thing I noticed in a comment from the conference that had to do with reaching the cap peak. My charge is a linear rise with no curvature. I am using 450volt caps and I think I should be using more like 50-75volt caps.

Anyway I think your right we need to get it running solely on the cap. So more testing is in order.
and Brent, thanks for bringing this up.
Les